Usefulness of aEEG in neonatal encephalopathies other than hypoxic ischemic encephalopathy has not been well studied and there is paucity of Indian data regarding the utility of aEEG monitoring in neonates. This study was planned to evaluate prognostic value of aEEG in term neonates with encephalopathy.

Encephalopathy was defined as subnormal alert state/altered neurological function which may be associated with seizures [5]. Etiology of encephalopathy was sub-grouped as hypoxic-ischemic encephalopathy (HIE), infective causes, transient metabolic causes, intracranial hemorrhage (ICH), dyselectrolytemia, and inborn errors of metabolism (IEM). A detailed neurological examination was performed, and level of alertness was defined as per the Volpe classification, and tone was assessed as per Amiel-Tison scale of tone assessment [6,7].

The aEEG monitoring was done by CFM Olympic Brainz Monitor (Natus Medical Inc.), with five biparietal hydrogel electrodes. Electrode impedance was main-tained below 10 ohms during monitoring. The aEEG was monitored for at least 24 hours for all enrolled neonates. A single investigator, trained for interpretation of aEEG findings in 10 aEEG recordings by a pediatric neurologist, noted the aEEG findings and scored as per the scoring used by Zhang, et al. [8].

Presence of any one of the following was defined as abnormal aEEG: any background activity other than continuous normal voltage, immature or absent sleep-wake cycle (SWC), and presence of any electrical seizure. Primary outcome was defined as subnormal level of alertness as per Volpe classification or any tone abnormality as per Amiel-Tison scale, at the time of discharge and/or death before discharge. Secondary outcomes were abnormal background activity, abnormal SWC, and electrical seizures, as defined above.

Statistical analysis: SPSS Software version 16 was used for data analysis. To determine the association between categorical variables, Chi square test was used as test of significance. P<0.05 was considered statistically significant. Diagnostic efficacy of aEEG background activity, aEEG sleep-wake cycle and aEEG seizures in predicting outcome at discharge was assessed by calculating specificity, sensitivity, positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio and negative likelihood ratio. Diagnostic efficacy of cumulative aEEG score in predicting outcome was assessed by ROC curve, suitable cut-off values were selected and AUC was calculated.　

Of all the term neonates with encephalopathy, 58 were finally enrolled in the study and followed-up till discharge or death (Fig. 1). The baseline characteristics are given in Table I. Out of these, 50 (86.2%) survived. aEEG findings were abnormal for 50 (86.2%) of enrolled neonates. Out of these, 31 (62%) had normal outcome and 19 (38%) had abnormal outcome at discharge, or died prior to discharge. There was a statistically significant association between abnormal aEEG findings and primary outcome. Abnormal aEEG had 100% sensitivity, 20.5% specificity, 38% PPV, 100% NPV, positive likelihood ratio 1.26 and negative likelihood ratio 0, to predict primary outcome (Web Table I).

Fig. 1 Study flow diagram.

Out of 58 neonates, 38 (65.5%) had abnormal back-ground activity. There was a statistically significant association between background activity and primary outcome (P=0.001). Out of 58 neonate, 19 (32.7%) had mature sleep-wake cycle and 39 (67.2%) neonates had immature or absent sleep-wake cycle, and there was a statistically significant association between sleep-wake cycle and primary outcome (P=0.001). A total of 43 (74.1%) neonates had electrical seizures, and aEEG seizures were significantly associated with primary outcome (P=0.002).

A cumulative aEEG score of 0-2 was seen in 32 (55%) neonates, and 26 (45%) had score >2. Out of 32 neonates with score of 0-2, 31 (97%) had normal outcome, ROC curve was plotted and cut-off value >2 was selected for high sensitivity and low false positive rate. There was a statistically significant association between cumulative aEEG score >2 and primary outcome (P=0.001) (Fig. 2).

Fig. 2 Receiver operator characteristic (ROC) curve of cumulative aEEG scoring to predict abnormal neurological outcome (AUC=0.746).

In this prospective observational study we assessed the characteristics of aEEG in 58 term neonates with encephalopathy and found that background activity, sleep-wake cycle and electrical seizures were significantly associated with primary outcome. In this study we also calculated cumulative aEEG scoring and we found that cumulative aEEG score of >2 was significantly associated with primary outcome. These results were largely consistent with previous studies.

In the present study we found that abnormal aEEG was significantly associated with abnormal outcome and it has high sensitivity, but low specificity to predict primary outcome. A meta-analysis by Chandrasekaran, et al. [9] showed similar results with pooled sensitivity of 87% and specificity of 36%. Similar to our findings, Van der Heide, et al. [10] noted significant association between aEEG background activity and neurologic outcome in neonates. Sewell, et al. [11] showed results opposite to our study with low sensitivity and high specificity. This may be due to inclusion of all grades of encephalopathy with various etiologies in our study, while they only included neonates with HIE. We found that aEEG cyclicity was significantly associated with primary outcome. Rhie, et al. [12] con-cluded in their study that delayed appearance of SWC was significantly associated with unfavorable neuroimaging in neonates with HIE, as was also seen in our study for all causes of encephalopathy.

Variane, et al. [13] showed that presence of recurrent aEEG seizures were associated with MRI brain abnormality and death, similar to this significant association with outcome was found in our study. Similar to this study, Luo, et al. [14] showed that aEEG scoring system has a higher specificity but low sensitivity as compared to individual components for abnormal outcomes.

Strengths of our study include enrolment of subjects with various causes of encephalopathy, albeit majority were HIE, and detailed study of the individual components of aEEG tracing and formulation of cumulative scoring cut-offs to predict short term neurological outcome.

Limitations of our study include a relatively small sample size, and inability to study long term neurological outcomes. Neonates admitted in the late stage of encephalopathy could have had different findings in aEEG if we had recorded aEEG at the onset of encephalopathy. Effect of ongoing drugs, and thera-peutic hypothermia were not taken into account, and the effects of postnatal age on aEEG findings were not studied.

To conclude, aEEG parameters such as abnormal background activity, absent sleep-wake cycling and presence of electrical seizures, either alone or in combination are associated with primary outcome of subnormal level of alertness or tone abnormality at discharge in term neonates with encephalopathy.

Note: Additional material related to this study is available with the online version at www.indianpediatrics.net

Ethical clearance: Institutional ethics committee, Kanchi Kamakoti CHILDS Trust Hospital Chennai; No. IEC-DNB/26/February2019, dated March 11, 2019.

Contributors: SGK: conceptualized the study, collected data, wrote the first draft of manuscript; NCK: study design, analysis, corrected manuscript and approved for final submission; HV: critical review of proposal, expert advice on data analysis and interpretation; SS: protocol development, supervising enrolment and outcome assessment; SSS: participated in planning of project and writing manuscript. All authors approved the final version of manuscript, and are accountable for all aspects related to the study.

Funding: None; Competing interest: None stated.

1. Glass HC, Kan J, Bonifacio SL, et al. Neonatal seizures: Treatment practices among term and preterm infants. Pediatric Neurol. 2012;46:111-5.

2. Hellström-Westas L, De Vries LS, Rosén I. An Atlas of Amplitude-integrated EEGs in the Newborn. CRC Press; 2008.

3. Shah DK, Boylan GB, Rennie JM. Monitoring of seizures in the newborn. Arch Dis Child Fetal Neonatal Ed. 2012;97:F65-9.

4. Hellström-Westas L, Rosén I, De Vries LS, et al. Amplitude-integrated EEG classification and interpretation in preterm and term infants. Neo Rev. 2006;7:e76-87.

5. D’Alton Mary E, Hankins GD, Berkowitz RL, et al, Neonatal encephalopathy and neurologic outcome. Pediarics. 2014;133;e1482.

6. Volpe JJ. Neurological examination: normal and abnormal features. In: Volpe’s Neurology of the Newborn, sixth edition, Elsevier. 2018;191-221.

7. Gosselin J, Gahagan S, Amiel Tison C. The Amiel Tison neurological assessment at term: Conceptual and methodological continuity in the course of follow up. Mental Retardation Dev Disab Res Rev. 2005;11:34-51.

8. Zhang D, Ding H, Liu L, et al. The prognostic value of amplitude-integrated EEG in full-term neonates with seizures. PLoS One. 2013;8:e78960.

9. Chandrasekaran M, Chaban B, Montaldo P, et al. Predictive value of amplitude-integrated EEG (aEEG) after rescue hypothermic neuroprotection for hypoxic ischemic encephalopathy: A meta-analysis. J Perinatol. 2017;37: 684.

10. Van der Heide MJ, Roze E, van der Veere CN, et al. Long-term neurological outcome of term-born children treated with two or more anti-epileptic drugs during the neonatal period. Early Human Dev. 2012;88:33-8.

11. Sewell E K, Vezina Gilbert, Chang Taeun, et al. Evolution of amplitude-integrated electroencephalogram as a predictor of outcome in term encephalopathic neonates receiving therapeutic hypothermia. Amer J Perinatol. 2018;35:277-85.

12. Rhie S, Chae KY, Jo HS, et al. Sleep-wake cycle on amplitude-integrated EEG and neuroimage outcomes in newborns. Italian J Pediatr. 2016;42:85.

13. Variane GF, Magalhaes M, Gasperine R, et al. Early amplitude-integrated electroencephalography for moni-toring neonates at high risk for brain injury. Jornal de Pediatria. 2017;93:460-6.

14. Luo F, Chen Z, Lin H, et al. Evaluation of cerebral function in high risk term infants by using a scoring system based on aEEG. Translat Pediatr. 2014;3:278.